LFP-Adaptive Dynamic Zoom-and-Tracking Dual-Band Neural Recording Front-End Using a Power-Efficient Incremental Delta-Sigma ADC

We report a local field potential (LFP) adaptive dynamic zoom-and-tracking front-end circuit for energy-efficient dual-band monitoring of subcortical neural signals. The proposed front-end employs an incremental delta-sigma ADC (IADC) to achieve multi-channel high-resolution neural recording with high energy-efficiency, through dynamically tracking slow-varying LFPs and zooming into a refined sub-LFP range to resolve action potentials (APs). The proposed zoom-and-tracking scheme can effectively reduce analog power consumption of the 2nd order IADC by 64.3% compared to the conventional design. In addition, the proposed architecture provides dual-band neural recording by reusing the tracked LFP signals. The front-end chip with four recording channels was implemented in 180 nm CMOS process. The IADC in the fabricated chip achieves the best FoMWalden (107.9 fJ/conv-step) and FoMSNDR (162.1 dB) among the state-of-the-art IADCs integrated in the neural recording front-ends. Also, the dual-band recording can reduce the post-processing hardware (FPGA) burden (> 53.8%) for band separation.